Time to first recurrence as a trial endpoint: time to change?

Area of Science:

• Cardiology
• Clinical Trials
• Biostatistics

Background:

• Interest in novel atrial fibrillation (AF) therapies is growing.
• Evaluating therapeutic efficacy typically requires large trials measuring outcomes like mortality or quality of life.
• Development of surrogate endpoints is crucial for reducing trial sample sizes.

Purpose of the Study:

• To assess if "time to first recurrence of symptomatic atrial fibrillation" is an appropriate surrogate endpoint for quality of life.
• To evaluate the validity of assumptions linking "time to first recurrence" to quality of life via episode frequency.

Main Methods:

• Review of existing assumptions regarding surrogate endpoints for AF management.
• Analysis of recent data from patients with implantable devices monitoring atrial tachyarrhythmias.
• Comparison of observed episode patterns with the Poisson distribution model.

Main Results:

• The assumption that "time to first recurrence" accurately reflects quality of life is questioned.
• Atrial tachyarrhythmia episodes, including symptomatic ones, tend to cluster, deviating from a Poisson distribution.
• Non-Poisson distributions necessitate larger sample sizes for detecting treatment differences in clinical trials.

Conclusions:

• "Time to first symptomatic recurrence" may not be a reliable surrogate endpoint for quality of life in AF management.
• Consideration of non-Poisson distributions is vital for designing future AF clinical trials.
• Alternative surrogate endpoints, such as episode frequency, severity, duration, or objective rate control measures, should be explored.